The northern transition area between Leonese and Castilian

Autor(en): Penny, Ralph

Objekttyp: Article

Zeitschrift: Revue de linguistique romane

Band (Jahr): 42 (1978)

Heft 165-166

PDF erstellt am: 03.10.2021

Persistenter Link: http://doi.org/10.5169/seals-399655

Nutzungsbedingungen Die ETH-Bibliothek ist Anbieterin der digitalisierten Zeitschriften. Sie besitzt keine Urheberrechte an den Inhalten der Zeitschriften. Die Rechte liegen in der Regel bei den Herausgebern. Die auf der Plattform e-periodica veröffentlichten Dokumente stehen für nicht-kommerzielle Zwecke in Lehre und Forschung sowie für die private Nutzung frei zur Verfügung. Einzelne Dateien oder Ausdrucke aus diesem Angebot können zusammen mit diesen Nutzungsbedingungen und den korrekten Herkunftsbezeichnungen weitergegeben werden. Das Veröffentlichen von Bildern in Print- und Online-Publikationen ist nur mit vorheriger Genehmigung der Rechteinhaber erlaubt. Die systematische Speicherung von Teilen des elektronischen Angebots auf anderen Servern bedarf ebenfalls des schriftlichen Einverständnisses der Rechteinhaber.

Haftungsausschluss Alle Angaben erfolgen ohne Gewähr für Vollständigkeit oder Richtigkeit. Es wird keine Haftung übernommen für Schäden durch die Verwendung von Informationen aus diesem Online-Angebot oder durch das Fehlen von Informationen. Dies gilt auch für Inhalte Dritter, die über dieses Angebot zugänglich sind.

Ein Dienst der ETH-Bibliothek ETH Zürich, Rämistrasse 101, 8092 Zürich, Schweiz, www.library.ethz.ch

http://www.e-periodica.ch THE NORTHERN TRANSITION AREA BETWEEN LEONESE AND CASTILIAN

In this article, the topic under discussion is that of dialectal and other linguistic boundaries. It is taken as axiomatic that such boundaries are rarely, if ever, sharp, but consist of a number of isoglosses, with more or less geographical space between each isogloss and the next. In establishing the linguistic boundaries even of the recent past, we are heavily dependent on knowledge of ethnic and political boundaries. In the case of the Leonese-Castilian boundary, the main difficulty is that not until the nth century is there a major political boundary separating the area where Leonese grew up from the area where Castilian originated. In previous periods, such political boundaries as existed between the two areas were of a minor nature and in any case their course is imperfectly known. By way of introduction, I shall review what is known of the ethnic and political boundaries in our area in seven historical periods : i) It has been claimed that the pre-Roman boundary separating the and the is one which has dialectal consequences *. Leonese certainly grew up in the territory of the Astures, while some have claimed that Castilian is the descendent of Cantabrian . However, the course of the Asturian-Cantabrian boundary is open to dispute, as we shall see. 2) Roman provincial boundaries often followed pre-existing tribal boundaries. For a short period (between 27 and 3 BC), immediately after the conquest of the hostile northern coastal area, Asturian territory was included within , while was added to the Citerior. However, from the beginning of the Christian era, Asturia (together with what is now and northern ) was also administered as part of the Citerior 2. Then, with Diocletian's administrative reform in the late

1. E.g. Menéndez Pidal, Orígenes del español, 5th ed., 1964, 214; R. Lapesa, Historia de la lengua española, 3rd ed., Madrid 1955, 28 ; A. Galmés de Fuentes, D. Catalán Menéndez Pidal, ' Un límite lingüístico ', RDTP, 2, 1946, 196-239 ; A. Zamora Vicente, Dialectología española, 2nd ed., Madrid 1967, 57. 2. C. Sánchez Albornoz, Orígenes de la nación española. Estudios críticos sobre la historia del reino de , I, 1972, 87. TRANSITION AREA BETWEEN LEONESE AND CASTILIAN 45

3rd century, both Cantabria and Asturia, together with Galicia and Portugal north of the Duero, were detached from the Citerior to form a new independent province of Gallaecia1. Therefore, Asturia and Cantabria belonged to different Roman provinces for only 24 years.

SantaneJU** OVlt«» T A U o t ft f V IXCAYA t »... '

'-.':'.-

L O M al 1 t ¦>- -

PALEGGIA B U, R Q OS

OUYQ05

3) The internal boundaries of the Roman provinces, the boundaries of the conventus, also often followed tribal boundaries. Indeed, the conventus, are most frequently defined in tribal terms — tribes A, B and C are said to belong to conventus X — so that we do not normally know the geographical extent of a conventus, unless we know the area occupied by given tribes. In this context, all we know about Asturia and Cantabria is that they belonged to different conventus, Asturia to the conventus asturicensis and Cantabria to the conventus cluniensis. 4) Diocesan boundaries are known often to follow Roman administrative boundaries (specifically, those of the conventus). But little precise is known about the diocesan boundaries of northern in the late Roman, Visi- gothic and early Reconquest periods. 5) Castile is of course part of the kingdom of León until the early nth century. From the mid 10th century we know reasonably precisely the boundary between the county of Castile and the neighbouring, less autonomous,

1. Sánchez Albornoz, Orígenes, 97. 46 R. PENNY counties to the West. It followed the Deva from the sea to the 1 and the from these mountains to the Duero 2. But this boundary appears to be a new creation ; it does not correspond with any known boundary of any previous period. 6) There is no difficulty in tracing the Leonese-Castilian political boundary from the nth to the 13th centuries. But the castilianization of eastern León (south of the mountains) was clearly going ahead rapidly at this time, so that the linguistic significance of this boundary (at least, of that part of it which lay south of the mountains) can only be small. 7) Lastly, the unification of León and Castile sooner or later removes any linguistic significance the political boundary south of the Cantabrian mountains may once have had.

The main purpose of this article is to discuss, first, how far the isoglosses which separate Leonese and Castilian are a reflection of the pre-Roman and Roman political boundaries in the area ; and, secondly, whether the medieval political boundaries have more or less linguistic consequences than the pre-Roman and Roman ones.

It has been claimed that Castilian is the descendent of the Latin of Cantabria, while Leonese is the descendent of that of Asturia. Menéndez Pidal's substratum account of the development of Latin F - hinges on this theory : he claims that the aspiration of F- is a feature of Cantabrian-influenced Latin, while the Latin of Asturia maintained f- intact3. We have seen that except for a few years after their conquest Asturia and Cantabria belonged to the same (although to different conventus). The precise geographical extent of the Asturian and Cantabrian areas is not known. Almost the sole concrete information on where their common boundary ran in the area between the sea and the Cantabrian mountains is a reference by Pomponius Mela (ist century AD) to a river salía as the boundary between the Astures and the Cantabri. This river salía may be identified with either of two rivers (although conceivably with neither) : the Sella, which rises in Riaño (NE León province), flows through the Peaks of Europe, through eastern Asturias and reaches the sea

1. R. Menéndez Pidal, Documentos lingüísticos de España, I, Reino de Castilla, Madrid 1919, 13. 2. Menéndez Pidal, Orígenes, 476. 3. Menéndez Pidal, Orígenes, 41. transition area between leonese and castilian 47 at Ribadesella. Alternatively, the salía may be identified with the Saja, some fifty miles to the East, which flows into the sea at (25 miles west of Santander). Of these two hydronyms, only Saja is a regular phonetic descendant of salía in its area (cf. palea > paja in Santander). In the area where the Sella flows, palea gives paya, so that from salía we would here expect *Saya. In the first edition of Orígenes del español, in 1926, Menéndez Pidal stresses that the f¡aspirate isogloss (whose existence he ascribes to ethnic factors) is located near the river Sella. However, he does not go as far as stating that this river had once constituted the boundary between Astures and Cantabri-. This step is taken by Sánchez Albornoz in 1929, in the process of attempting to establish the geographical areas occupied by each of the pre-Roman tribes of northern Spain 1. Sánchez Albornoz, then, identifies the river salía, mentioned by Pomponius Mela, with the modern Sella. But in ignoring the Saja and adopting the Sella, Sánchez Albornoz leans almost exclusively on the linguistic evidence presented by Menéndez Pidal in Orígenes1 (the fjaspirate isogloss). Sánchez Albornoz supposes that the form salía has been the victim of corrupt transmission and that the original spelling was *saelia. This neatly, if somewhat arbitrarily, takes care of the tonic vowel of sella, but fails to resolve the difficulty posed by the l + yod cluster. The dialects of eastern Asturias do not confuse the reflex of L + yod ([y]) with that of -ll- ([1]) ; and since yeísmo is unknown in this area, there is no possibility that the form Sella has been arrived at through hypercorrection (in the way that medieval Castilian Mayorca became Mallorca). Even starting from *saelia, Sánchez Albornoz's argument will not account for the form Sella, but only for *Seya. In spite of these phonetic difficulties, the argument comes full circle with the appearance in 1950 of the third edition of Orígenes del español, in which Menéndez Pidal uses Sánchez Albornoz's conclusions on the geographical extend of Cantabria in support of his substratum explanation of the aspiration of F-2. The linguist and the historian each support the other, but if we once deny the historical basis of their argument (the identity of salía and Sella), the argument becomes untenable. In further support of the idea that the Sella represents the western boundary of the Cantabri, Menéndez Pidal uses the place-name Cofiñal. This is

1. C. Sánchez Albornoz, Boletín de la Real Academia de la Historia, 45, 1929, 3T5"95 Orígenes, 51-100 ; see p. 56). 2. See p. 214. 48 R. PENNY a locality in Riaño (northeastern León province), not far to the West of the upper Sella. The etymon of Cofiñal is conceivably *confiniale, and such a name might be expected to occur near a tribal or other political boundary. Menéndez Pidal claims that Cofiñal (with its -f- preserved) marks the eastward limit of the territory of the Astures and therefore the boundary between Astures and Cantabri. However, Menéndez Pidal mentions, briefly and in passing, other toponyms whose etyma contain fines (or a derivative) and among these a Cohiño in Santander. In fact, the accompanying map shows two localities with this name in the province. One of these lies on the river Besaya, a tributary of the Saja. Cohiño is at least as likely as Cofiñal to mark the ethnic boundary with which we are concerned ; however, there is no convincing reason to believe that either Cohiño or Cofiñal marks the Asturian-Cantabrian boundary, since later and more minor boundaries could be responsible for the names. I conclude from this discussion that the river referred to by Mela as the salía is the Saja and that if this river in fact marked the western limit of the Cantabri proper, then the apparent ' aspiration ' of Latin F- has nothing exclusively to do with the Cantabri, since this feature stretches as far West as the Sella. The reasons for the position of the //aspirate isogloss along the Sella are as impenetrable as those for almost any other peninsular isogloss outside the areas of Christian reconquest. There is considerable amount of evidence to show that the modern distribution of dialect features does not depend on the relative disposition of the Astures and the Cantabri (or of other tribal groups). I shall summarize the evidence : 1) The name salía should be identified, as we have seen, with the Saja and not with the Sella. 2) The area between the Deva and the Ría de Santander (the western half of the modern province) was named in the Middle Ages the Asturias de . The explanation is probably the obvious one : this area, like the Asturias de Oviedo, which stretches up to the Deva, had previously been territory belonging to the Astures and not to the Cantabri. 3) The ethnographical evidence presented by Krüger x and Caro Baroja 2 stresses the cultural unity of the Astures and the Cantabri, with each other and with other northern tribal groups.

1. F. Krüger, El léxico rural del noroeste ibérico, Madrid 1947. 2. J. Caro Baroja, Los pueblos del norte de la peninsula ibérica, 2nd ed., San Sebastián 1973, especially p. 258-63. TRANSITION AREA BETWEEN LEONESE AND CASTILIAN 49

4) Almost without exception, the earliest notarial documents which show replacement of Latin f- were not written in the area traditionally acknowledged to be Cantabrian. One was written in Nájera (territory of the ) another in (territory of the Turmogi or Murbogi). Only the Santoña documents were written in Cantabria, even if at the far eastern extremity. If this discussion has so far been negative, I have on another occasion shown that it is possible to explain the different Hispanic results of F- along phonetic/phonological lines 1. To summarize that previous argument, it is widely thought that early Latin F was a bilabial ([cp]) ; if this sound were preserved in the Latin of Spain, it is possible that [cp] became [h] before back vowels through dissimilation of bilabial features ; the alternation between [

I shall now turn to the medieval political boundaries separating areas of Leonese speech from those of Castilian speech. Only in the early ioth century was the power of the counts of Castile, based in what is now northern , extended north into Santander. This in spite of the fact that the population of the primitive Castile originated north of the Cantabrian mountains and individuals owned land on both the northern and southern sides 2. In about 950, Fernán González had become count not only of Burgos, but also of Castilla Vieja, Cerezo, Lantarón, and the Asturias de Santillana, that is, the provinces of Burgos, Santander (except Liébana in the Southwest), Álava, Vizcaya and Guipúzcoa as far East as San Sebastián. In the time of Sancho García (955-1017), the grandson of Fernán González, we know specifically that the boundary of the expanded Castile with León proper followed the Deva and the Pisuerga 3. In this period, Liébana has close links with the counties of Saldaña and Carrion and is not part of Castile. In 1030, the counties of Liébana, Saldaña, Carrión and Monzón (i. e. what is now southwestern Santander and the province of ) were incorporated into Castile. The western frontier of Castile was thus moved

1. ' The re-emergence of /f / as a phoneme of Spanish', ZRP, 88, 1972, 463-82. 2. J. Pérez de Urbel, El condado de Castilla, I, Madrid 1969, 267. 3. The Deva today forms the boundary between Asturias and Santander in the coastal region. Revue de linguistique romane. 4 50 R. PENNY

from the Pisuerga to the Cea, although in the far North the lower course of the Deva continued to mark the boundary 1. We can now ask what influence on dialectal distribution is exercised by the medieval boundary which followed the Deva and by the preRoman boundary which followed either the Sella or, as I have argued, the Saja. We have seen that the Sella coincides fairly well with the f¡aspirate isogloss (W [farina] vs E [harina]) 2. Another isogloss also runs in this area, although its precise course is not well established : that which separates W ome, fame, etc., from E ombre, hambre, etc. According to ALPI 3, the Deva coincides, at least in its lower reaches, with a number of isoglosses : i) that which separates W [y] in [abéya], [agúya], [6éya], from E [h] in [abéha], [agúha], [0éha], abeja, aguja, ceja ; 2) that which separates W [y] in [kuyár] from E [e] in [kucár], cuchara ; 3) that which separates W [s] in [kása], [kósu], [ése], from E [h] in [káha], [kóhu], [éhe], caja, cojo, eje ; 4) West of the Deva the reflex of bove is [gwé:], [gwë:], while East of the Deva its reflex is [gwéi], buey ; 5) West of the Deva, VL camminu gives [kamïrj], while East of the river the result is [kaminu], camino ; 6) West of the Deva, crine gives iklina], while on the eastern side it gives [klírj], crin. According to my own investigations 4, the river Saja coincides (in a very approximate way) with the following three isoglosses : 1) West of the Saja, the reflex of *ambosta is [amboOá :], etc., with preservation, of -mb-, while to the East of the Saja the reflex is [mo0á], ambuesta However, [lamber] (< lamberé) and [kambéra] (< camba + -aria), camino carretero, with -mb- preserved, are used throughout the province. 2) West of the Saja, the 3rd sing. près. ind. forms of -er¡-ir verbs lack ¡-ef : [há0], [á0], etc., while to the East of the Saja /-e/ is present : [há6a], [ä6s], etc., hace. 3) The dialects West of the Saja are loista (retaining the etymological

1. Pérez de Urbel, Condado de Castilla, III, 236. 2. Cf. Galmés de Fuentes, Catalán Menéndez Pidal, ' Un límite lingüístico '. 3. Atlas lingüístico de la peninsula ibérica, I, Fonética i, Madrid 1962, maps 6, !2- 41." 57 ; 32. 50, 75 ; 28 ; 33 ; 53. 4. Findings of field-work to be published shortly. transition area between leonese and castilian 51 functions of lo and le), while the dialects East of the Saja are teísta (using le as a direct object pronoun both for persons and things). But cutting across these possible boundaries there are a number of features which I have found to show continuity from western or central Asturias right through to eastern Santander l : 1) Vocalic metaphony is strongest in central Asturias and in central Santander, but in a weaker form is to be observed also in eastern Asturias and western Santander. 2) Aspirate descendants of f- (+ ó) have a similar distribution ; e. g. [hwénte]/[Avwénte]/(< fónte), fuente. 3) Palatalization of L- is observable from western Asturias to eastern Santander, at least in words like [lár]/[yár], hogar (< lare), although it is true that in other words l- is represented by [1] in Santander (e. g. [lamber], lamer, [limjágu], limaza). 4) Palatalization of N- is observable over the same area in some words ([núctu], nudo, etc.). 5) Pronoun and adjective agreement with a mass-noun (of either gender) is in /-0/ at least from central Asturias to central Santander (e. g. la hierba hay que [segalo], la harina está muy [blanko]) 2. The conclusion to which this article is that the possible pre-Roman boundaries between Astures and Cantabri have little (if any) influence on the modern distribution of isoglosses separating Leonese dialects from Castilian. Two isoglosses happen to run near the Sella, but we have seen that this river is unlikely, in fact, to have formed the boundary between the Astures and the Cantabri. The Saja coincides in only an approximate way with three isoglosses and, with the exception of the [mb]/[m] isogloss, the innovations concerned (loss of /-e/ in verbs and leísmo) postdate by many centuries the implantation of Latin in this area. The Castilian-Leonese boundary of 1030 is much more important. The river Deva coincides with a number of isoglosses reflecting Castilian innovations of appropriate antiquity. But in the most important body of features examined, continuity can be observed from central Asturias (sometimes from western Asturias and beyond) right through to eastern Santander. This continuity reflects the

1. Findings of field-work to be published shortly. 2. For a discussion of this phenomenon and full bibliography, cf. R. J. Penny, ' Mass-nouns and metaphony in the dialects of Northwestern Spain ', Archivum Linguisticum, n. s. 1, 1970, 21-30. 52 R. PENNY pre-ioth-century political and cultural situation, one of cultural and linguistic similarity along the north coast, embracing Asturias and Santander (and perhaps other areas like Galicia). Innovations arising in northern Burgos begin to spread into Santander from the ioth century, but do not generally pass beyond the Deva, which marks the political frontier from about 950. After the union of León and Castile in the 13th century (when the Deva boundary ceases to be one of any great political importance), eastern Asturias resists castilianization (at least until modern times) since this area was always sufficiently remote (and was always a source of emigration rather than a goal for immigrants) for its linguistic character to remain unchanged. That is to say, isoglosses reflecting innovations which originate in northern Burgos have not progressed beyond the Deva l.

Westfield College, University of London. Ralph Penny.

1. This article is a re-working of a paper read to the annual meeting of the Association of Hispanists of Great Britain and , Glasgow, 1975. I am grateful to my colleague Professor Alan Deyermond for reading the typescript and suggesting many improvements.